Wind Sound Engine

This idea for an outdoor installation questions whether the interdisciplinary obsession with the scientific performativity of data as art (data that may at a first glance seem chaotic to the observing artist) may not in fact be a spontaneous desire to render this data back into chaos. That is, to take something that is organized according to a post-human logic and treat it as a source of random numbers with which to trigger audio-visual events imitating the spontaneity of art; rendering the chaotic meaningful while simultaneously revealing it as, ultimately, chaotic. Or perhaps making what is meaningless in abundance meaningful in simpler ways that, nevertheless, remain profoundly disconnected in meaning, following Orit Halpern’s thought in Beautiful Data.

The democratization of audiovisual technology for mass consumption allowed the many to do what was formerly given only to a few. This process implied a re-thinking of notions of high vs low art, authorship, copyright. The reduction and disappearance of Benjaminian aura is a result of this process, that which is sometimes referred to as ‘divine inspiration’, or more recently as ‘human creative agency.’

Aeolus at The Eden Project

Suppose we were to take this process to its logical extreme and remove human agency altogether, what would art look like generated entirely by non-human factors? Chris Chafe, a computer music and acoustics researcher, does just that using publicly available data sets: “everything from CO2 levels to Gross Domestic Product data.” Another implementation is Brian “Data-Driven DJ” Foo’s sonification of the income gap on the NYC subway.

Perhaps this is the sort of thing that interests Douglas Eck, research scientist at Google who wants to see the TensorFlow library for machine intelligence used for “music and art generation using deep neural networks,“ though anticipating criticism from the arts and humanities areas, he makes sure to mention that his project “[is] not about replacing musicians and artists with machines, but rather about understanding core aspects of creativity such as surprise and attention.” In this comfortably sponsored quest, Eck has frequently been joined by “brain on music” Dan Levitin as the two have attempted to unravel the mysteries of human preferences and machine recommendations. The ethical implications of corporation-directed research aside (NB I am not against corporate sponsorship, but against corporate guidance and management of research that twists a philanthropic donation into underpaid corporate outsourcing), such endeavors still involve human participation.

Taking things then one step further, I envision data driven compositions which derive their information from environmental sources, from chaos. Is there a semiotics of sound ecology? While language and by extension, data are interconnected systems within which signs are contextualized, a sound needs no interpreter – it just is. Sound ecologists strive to isolate sounds, to classify them, and thus to turn sound into a language. Furthermore, sound–whether spontaneous or reproduced–is always something that has already happened. The idea is then to preserve the potentiality of sound through sound sculpture installations which can be performed by a post-Anthropocene environment. See for example Luke Jerram’s wind-powered “Aeolus“, Gisèle Trudel’s Ælab and Grupmuv, David Letellier and Herman Kolgen’s Eotone.


…if we were to define the semi-autonomous doxa of information strictly on the premise that information is something we consume, like news or entertainment, then we might say that our current fundamental belief is that all useful knowledge (what Bourdieu calls “practical sense”) simply coincides with the “objective” world as that world comes to us mediated through information technologies whose real-time, interactive quality neutralizes the sense of mediation and so constitutes the meaning of objectivity. (The sense of being “live” on today’s news broadcasts, for example, is conveyed by such unedited, real-time effects as the percussion of wind upon a journalist’s microphone—a phenomenon that has no natural counterpart and thus amounts to a technology of natural reality. “Real-time” or “live”—no matter how mediated—simply is the same as “real” or “true.”) Information consumed without concern for technological mediation, we may thus say, is our contemporary habitus. It is the habitual information environment in which “subjective principles of organization” (as Bourdieu puts it) are deeply in-formed by a world defined as technology-object. If there is also a strong sense of dissonance in such habitus (the conviction that IT overwhelms us with a “data smog” of mere facts that don’t make sense), then such dissonance is subsumed within what could be called the general consonance of dissonance (the “expected” meaninglessness of the contemporary world). (Liu 41)

– from Liu, Alan. The Laws of Cool: Knowledge Work and the Culture of Information. 1 edition. Chicago: University Of Chicago Press, 2004.


A possble platform where such research can be seen and submitted is the annual International Workshop on Musical Metacreation (MUME), this year to be held in Paris. From the CFP:

MUME aims to bring together artists, practitioners, and researchers interested in developing systems that autonomously (or interactively) recognize, learn, represent, compose, generate, complete, accompany, or interpret music. As such, we welcome contributions to the theory or practice of generative music systems and their applications in new media, digital art, and entertainment at large.


Aeolian Harp